Datasets:

ronenEl
/

Policy_Llama3.2-1B_PRM_Llama3.1-8B-best_of_n-completions

Modalities:

Formats:

Size:

Libraries:

Dataset Viewer

Subset (5)

MATH-500--T-0.8--top_p-1.0--n-256--max_tokens-2048--bsz-8--seed-0--agg_strategy-last--evals · 9 rows

Split (1)

train · 9 rows

Search is not available for this dataset

n int64 1 256	acc_naive float64 26.8 45.2	acc_weighted float64 26.8 50	acc_maj float64 26.8 45.2
2	29.6	29.6	26.8
32	42	46	44.6
256	45.2	50	45.2
64	43.2	47.2	44.6
128	44.6	49.4	45
16	42.2	43.4	41
8	40.4	39.8	35.6
1	26.8	26.8	26.8
4	35.6	34.6	30.6

README.md exists but content is empty.